lable at ScienceDirect

Solid State Sciences 13 (2011) 981e992

Contents lists avai

Solid State Sciences

journal homepage: www.elsevier .com/locate/ssscie

Study on the preparation and properties of silver-doped phosphateantibacterial glasses (Part I)

A.A. Ahmed a, A.A. Ali a,*, Doaa A.R. Mahmoud b, A.M. El-Fiqi a

aGlass Research Department, National Research Centre, Dokki, Cairo, EgyptbNatural and Microbial Products Laboratory, National Research Centre, Dokki, Cairo, Egypt

a r t i c l e i n f o

Article history:Received 14 October 2010Received in revised form26 January 2011Accepted 11 February 2011Available online 19 February 2011

Keywords:Antibacterial glassesSilver-doped phosphate-based glassesGlass dissolution

* Corresponding author.E-mail address: ali_nrc@hotmail.com (A.A. Ali).

1293-2558/$ e see front matter � 2011 Elsevier Masdoi:10.1016/j.solidstatesciences.2011.02.004

a b s t r a c t

Silver-doped phosphate antibacterial glasses were prepared by the melting method. The antibacterialeffects of some undoped and silveredoped glasses of compositions 65P2O5e10CaOe(25ex) Na2O,70P2O5e20CaOe(10ex) Na2Oand (70ex) P2O5e30CaO, (where x¼ 0, 0.5, 1.2 Ag2O), against Staphylococcusaureus, Pseudomonas aeruginosa and Escherichia coli micro-organisms using agar disk-diffusion assayswere investigated. The structures of some glasses were studied by XRD, FT-IR, and UVeVIS spectroscopy.The variation of pH with dissolution rate was studied. The tested silver-free and silver-doped glassesdemonstrated different antibacterial effects against the tested micro-organisms. For silver-free glasses, anincrease in inhibition zone diameter (zone of no bacterial growth) was seenwith the decrease inwater pH.Silver-doped glasses showed an increase in inhibition zone diameter with increasing Ag2O content. Thelow pH produced by glass dissolution was certainly a critical factor for glass antibacterial effect. The morethe phosphate ions released the lower is the pH and the greater the antibacterial effect.

� 2011 Elsevier Masson SAS. All rights reserved.

1. Introduction

Glasses in the P2O5eCaOeNa2O system have a chemicalcomposition similar to that of the inorganic phase of bone. Theseglasses consist of PO4 tetrahedra, which can be attached toa maximum of three neighboring tetrahedra forming a threedimensional network as in vitreous P2O5 [1]. Addingmetal oxides tothe glass leads to a depolymerization of the network, with thebreaking of PeOeP linkages and the creation of non-bridgingoxygens. The modifying cations can provide ionic cross-linkingbetween the non-bridging oxygens of two phosphate chains, thusincreasing the bond strength of this ionic cross-link and improvingthe mechanical strength and chemical durability of the glasses [2].Thesephosphate-basedglasses are aunique class ofmaterials in thatthey are completely degradable; whereas silica-based glasses arerelatively stable to hydrolysis. Furthermore, the degradation ofphosphate-based glasses can be tailored to suit the end applicationand the rate atwhich they hydrolyze canvary quite considerably [3].

Various types of silver-doped inorganic antibacterial materialshave been developed e.g. zeolites, calcium phosphate, silica gel, andborosilicate glass and some of them are now in commercial use.

son SAS. All rights reserved.

Silver ions are effective against a broad range of micro-organismsincluding G- bacteria e.g., Pseudomonas aeruginosa, yeast e.g.,Candida albicans, and Gþ bacteria e.g. Staphylococcus aureus [4,5].Therefore, silver ions have been commercially used to takeadvantage of its antibacterial properties e.g. silver nitrate, colloidalsilver, and certain other silver compounds are among the mostgenerally used bactericidal agents. A large number of healthcareproducts now contain silver ions, principally due to its low toxicityto human cells and high antibacterial effect. Such products includesilver-coated catheters, and wound dressings [6]. Phosphate-basedglasses are materials of technological importance due to theirsuperior physical properties compared to silicate glasses e.g., lowmelting temperatures, low glass transition and low softeningtemperatures, and high thermal expansion coefficients [7,8]. ThusPBGs can be prepared and processed easily at lower temperatures.In addition, phosphate-based glasses enjoy a range of composi-tional and structural possibilities (ultra, meta, pyro, and ortho) thatfacilitate tailoring chemical and physical properties of interest forspecific technological applications. Controlled-release glasses(CRGs) were first developed in the 1970s primarily for use in foodproduction industries [9]. Drake and Allen [10] found that PBGswith a suitable composition would dissolve in water with zero-order rate constant, and by controlling the composition it waspossible to produce glasses which would completely degrade inwater from hours to years thus can, over a prolonged period, release

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992982

any additional constituents incorporated into them. Hence it hasbeen possible to use pellets of CRGs containing metal ions such ascopper and cobalt, as pesticides, fungicides and in animal feeds.CRGs are manufactured in a similar way to conventional soda-limesilica glasses in that the constituents are heated to temperaturesabove 1000 �C, then cast into various forms such as solid blocks,powder, granules, tubes, fiber or wool. The incorporation of well-known silver, copper or zinc antibacterial metal ions in several glasssystems has a proven negative influence on the growth of bacteriaand fungi [11,12]. Where in presence of an aqueous medium ormoisture, the glass will gradually dissolve and at the same time,silver, copper, or zinc ions are released during its dissolution toprovide an antibacterial effect. Generally, antibacterial glasses canbe manufactured either by addition of an antibacterial agent to theglass batch prior to their manufacture or by post-treatmentprocesses e.g. ion-exchange or surface coating. This work is anattempting to prepare and study the antibacterial effect of highdissolution silver-doped phosphate glasses.

2. Experimental

2.1. Glass preparation

All batches were prepared from chemically pure grade chem-icals in the powder form. P2O5 was introduced as (NH4H2PO4)(99.0% Merck), Calcium oxide (CaO) as Calcium carbonate(CaCO3) (99.5% SRL) Sodium oxide (Na2O) as sodiumcarbonate (Na2CO3) (99.5% s.d. fine-chem), and silver oxide (Ag2O)as silver nitrate (AgNO3) (99.9% SRL).The appropriate amounts ofthe starting materials of each batch equivalent to 50 g glass wereaccurately weighed, thoroughly mixed and then transferred toporcelain crucibles. Before melting, the batches were calcinedslowly in an electric muffle furnace at a temperature in the range of350e550 �C in order to get rid of the gaseous decompositionproducts of the batch materials, e.g. H2O, NH3, NO2, and CO2 and tominimize the evaporation tendency of P2O5. Calcination wascontinued until the decomposition of the batch materials andevolution of gaseous products came to an end. All the batches weremelted in disposable porcelain crucibles inside an electricallyheated furnace in the range 800e1200 �C. The melting time wascontinued for 1 h to 2 h depending upon the chemical composition.During melting, the melt was stirred manually by swirling aboutseveral times to ensure homogeneity and to get ride of gas bubbles.The melt was then cast on a preheated stainless steel plate in theform of rectangular slabs which subsequently annealed in a mufflefurnace maintained at a temperature in the range 200e450 �C for20 min. The muffle furnace was then switched off and the glasssamples were left overnight to cool slowly to room temperature.The visible characteristics e.g. color; transparency, and homoge-neity, of all samples prepared in this work were investigated usingthe normal visual observations.

2.2. XRD measurements

To ensure the glassy state, some selected samples were char-acterized with powder X-ray diffraction technique which iscommonly used to verify the amorphous state of glassy materials.In the XRD spectra of glassy materials a halo is seen instead ofdiffraction peaks. the samples were finely ground in an agatemortar and X-ray diffraction spectra were obtained using a BrukerD8 Advance X-ray diffractometer at room temperature with Ni-filtered Cu Ka radiation (l ¼ 0.15418 nm), generated at 40 kV and40 mA. Scans were performed with a step size of 0.02� and a steptime of 0.4 s over an angular range 2q from 4� to 70�.

2.3. FT-IR absorption measurements

The FT-IR absorption spectra of some selected undoped andsilver-doped glasses (0, 0.5, 1 and 2 mol% Ag2O) were recorded atroom temperature in the frequency range 400e4000 cm�1 using aninfrared spectrometer (Jasco FT-IR 6100). The measurements weremade by the KBr disc technique in which discs were prepared bymixing and grinding a small amount of glass powder with spec-troscopic grade anhydrous KBr powder and then pressed undervacuum and pressure of 6 ton/cm2 into clear disks (1.2 cm indiameter and about 0.5 mm in thickness). All measurements wererecorded with a resolution of 4 cm�1.

2.4. UVeVIS absorption measurements

UVeVIS absorption spectra were measured for some undopedand silver-doped glasses (0, 0.5, 1 and 2 mol% Ag2O). Polished glasssamples having dimensions 3 cm � 1 cm and of the same thickness(2 mm) were scanned in the range from 200 to 1000 nm usinga UVeVIS spectrometer (T80þ, PG instruments Ltd.).

2.5. pH measurements

The pH changes of the distilled water during the dissolution ofsome undoped and silver-doped glasses were measured at everyhour and up to 6 h using IQ 140 pH-meter (IQ Inc. USA). The pHelectrode was calibrated using pH calibration standards (ColourkeyBuffer Solutions BDH, UK).

2.6. Antibacterial activity test

The antibacterial activities of undoped and silver-dopedP2O5eCaOeNa2O glasses were tested against bacterial species ofAmerican Type Culture Collection (ATCC); S. aureus (ATCC, 25923),E. coli (ATCC, 25922), and P. aeruginosa (ATCC, 27853) using the agardisk-diffusion assays.

3. Results

3.1. Glass forming region (GFR)

The glass forming regions and the compositions prepared in thesystems P2O5eCaOeNa2Oex Ag2O and P2O5eCaO-x Ag2O, x¼ 0.5, 1and 2 mol % are illustrated in Figs. 1e3. Clear circles denotehomogeneous, transparent and colorless glasses as confirmed byXRD. Black circles denote samples that showed metallic silverparticles precipitation. Fig. 1 and Fig. 2 showed that the composi-tions containing � 60 mol% of P2O5 in the quaternary systemP2O5eCaOeNa2Oex Ag2O, x ¼ 0.5 and 1 mol%, formed homoge-neous, transparent and colorless silver-doped glasses, whereas itwas not possible to obtain homogeneous glasses for the composi-tions containing � 55 mol% of P2O5 since these compositionsshowed precipitations of metallic silver particles. Also it can beseen from Fig. 3 that the compositions containing � 65 mol% ofP2O5 in the quaternary system P2O5eCaOeNa2Oe2Ag2O formedhomogeneous, transparent and colorless silver-doped glasses.Nevertheless, among compositions containing 60 mol% of P2O5,only three compositions which contain 10, 15 and 20 mol% of Na2Oformed homogeneous, transparent and colorless silver-dopedglasses. For other compositions containing 5, 25, 30 and 35 mol %Na2O, homogenous silver-doped glasses could not be obtainedsince these compositions showed precipitations of metallic silverparticles. Overall, A glass forming region containing � 60 mol% ofP2O5 was observed in the quaternary system P2O5eCaOeNa2OexAg2O, x ¼ 0.5, 1 and 2 mol%.

Na2

O + 2 Ag2

O0 5 10 15 20 25 30 35 40 45 50

P2

O5

50556065707580859095100

CaO

05

101520253035404550

A

B

C

D

E

F

G

H

I

JK

1234567891011

.

Mol %

...

2 mol % Ag2

O

Fig. 3. Ternary phase diagram showing the GFR and the glass forming compositions inthe quaternary system P2O5eCaOeNa2Oe2Ag2O. B: Homogeneous glass (Transparentand Colorless). C: Metallic silver particles precipitation.

Na2O + 0.5 Ag

2O0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

P2O

5

3035404550556065707580859095100

CaO

05

10152025303540455055606570

A

B

C

D

E

FG

H

I

JK

LM

N

123456789101112131415

...

..... .....

Mol %

0.5 mol% Ag2O

Fig. 1. Ternary phase diagram showing the GFR and the glass forming compositions inthe quaternary system P2O5eCaOeNa2Oe0.5 Ag2O.

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992 983

X-ray diffraction measurements were performed on someP2O5eCaOeNa2OeAg2O glasses containing 2 mol % Ag2O. The X-raydiffraction patterns obtained for these glasses are displayed inFig. 4. As shown in this figure, the X-ray diffraction patterns showedno sharp peaks thus indicating the absence of formation of anycrystalline phases and ensuring the amorphous nature of theseprepared samples.

3.2. FT-IR absorption spectroscopy

FT-IR absorption spectroscopy was used to detect any change inthe structure of some P2O5eCaOeNa2O and P2O5eCaO glasses,Table 1, that may have occurred as a result of introducing Ag2O intothese glasses and to obtain essential information concerning thearrangement of the phosphate structural units in the phosphateglass network. As shown in Fig. 5a, b and c no newabsorption bandswere detected on the addition of x Ag2O, x ¼ 0.5, 1 and 2 mol%to the P2O5eCaOeNa2O base glasses. Also, it can be seen that mostof the bands observed in the base glasses do not show significantshifts with the Ag2O addition. Most of bands appeared on the same

Na2

O + 1 Ag2

O0 5 10 15 20 25 30 35 40 45 50

P2

O5

50556065707580859095100

CaO

05

101520253035404550

A

B

C

D

E

FG

H

I

JK

1234567891011Mol %

1 mol % Ag2

O

Fig. 2. Ternary phase diagram showing the GFR and the glass forming compositions inthe quaternary system P2O5eCaOeNa2Oe1 Ag2O.

position, whereas few bands shifted to slightly higher frequencies.It can also be seen that the intensities of some bands increased withthe increase in the Ag2O content. Sixmain absorption bands locatedin the regions at about 445e472 cm�1, 720e750 cm�1,907e918 cm�1, 1024e1050 cm�1, 1100e1109 cm�1 and1298e1332 cm�1 along with two shoulders at about 530 cm�1 and1170 cm�1 were observed in all FT-IR spectra of P2O5eCaOeNa2Oglasses, Table 2.

3.3. UVeVIS absorption spectra

Fig. 6a, b and c shows the UVeVIS absorption spectra recordedin the wavelength range 200e1000 nm at room temperature for70P2O5e20CaOe(10ex) Na2Oex Ag2O, (70ex) P2O5e30CaO-x Ag2Oand 65P2O5e10CaOe(25ex) Na2O-x Ag2O glasses, x ¼ 0, 0.5, 1 and2mol%. From this figure, it can be seen that the undoped and silver-doped glasses reveal no absorption peaks in the visible region. Twoabsorption peaks were observed in the ultraviolet region, one atabout 210 nm was observed for all undoped glasses. A small red

Fig. 4. The XRD patterns for some P2O5eCaOeNa2OeAg2O glasses containing 2 mol %Ag2O.

400 600 800 1000 1200 1400 1600 1800 2000

1644

Abs

orba

nce

(a.u

.)

x = 0

x = 2

x = 1

129810961170

1017911735

530450 x = 0.5

400 600 800 1000 1200 1400 1600 1800 2000

x = 1

x = 0

Abs

orba

nce

(a.u

.)

Wavenumber (cm-1)

wavenumber (cm-1)

wavenumber (cm-1)

1644133211701109

1024917743

528445

x = 0.5

x = 2

400 600 800 1000 1200 1400 1600 1800 2000

x = 2

11701100 1300 1644720

Abs

orba

nce

(a.u

.)

x = 0

1024907530448

x = 0.5

x = 1

a

b

c

Fig. 5. FT-IR absorption spectra for (a): 70P2O5e20CaOe(10ex) Na2Oex Ag2O,(b):(70ex) P2O5e30CaO-x Ag2O and (c) :65P2O5e10CaOe(25ex) Na2O-x Ag2O glasses,x ¼ 0, 0.5, 1 and 2 mol%.

Fig. 6. UVeVIS absorption spectra (a): 70P2O5e20CaOe(10ex) Na2Oex Ag2O,(b):(70ex) P2O5e30CaO-x Ag2O and (c) :65P2O5e10CaOe(25ex) Na2O-x Ag2O glasses,x ¼ 0, 0.5, 1 and 2 mol%. The inset shows the region 200e320 nm.

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992984

shift (a shift to longer wavelength) in the position of this absorptionband was observed (in all UV-spectra of Ag2O-doped glasses) withincreasing Ag2O content and strong absorption peak at about230 nm was observed for all silver-doped glasses.

3.4. pH measurements and dissolution rates

The pH changes and dissolution rates measured during thedissolution of some undoped and Ag2O-doped P2O5eCaOeNa2Oand P2O5eCaO glasses in distilled water at 37 �C for different timeintervals up to 6 h are listed in Tables 1 and 3 and Fig. 7. As shown inTable 3 and Fig. 7 a fast drop in pH of distilled water (w5.5) wasseen through the first hour of glass dissolution and then the pHdecreased slowly with increasing time of dissolution. The glasswith the highest dissolution rate, H3, shows a higher decrease in pH

Dissolution Time ( Hours)0 1 2 3 4 5 6 7

Hp

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

G5

G5Ag0.5

G5Ag1

G5Ag2

Dissolution Time ( Hours)0 1 2 3 4 5 6 7

Hp

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

G7

G7Ag0.5

G7Ag1

G7Ag2

Dissolution Time ( Hours)0 1 2 3 4 5 6 7

Hp

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

H3

H3Ag0.5

H3Ag1

H3Ag2

a b

c

Fig. 7. pH variation with time during the dissolution of (a): 70P2O5e20CaOe(10ex) Na2Oex Ag2O, (b):( 70ex) P2O5e30CaO-x Ag2O and (c):65P2O5e10CaOe(25ex) Na2O-x Ag2Oglasses, x ¼ 0, 0.5, 1 and 2 mol%.

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992 985

than that showed by the glasses having lower dissolution rates.That means the pH is dependant on the glass dissolution rate.

3.5. Antibacterial activity

The antibacterial effects of undoped and2 mol % Ag2O-dopedP2O5eCaOeNa2O and P2O5eCaO glasses were tested in vitro againstS. aureus as Gþ, P. aeruginosa and E. coli as G- micro-organisms usingagar disk-diffusion assays. The results of agar disk-diffusion assaysconducted for 24 h at 37 �C are shown in Figs. 8 and 9. The anti-bacterial activity of the glass was confirmed by the presence of aninhibitory zone (i.e. zone of no bacterial growth) around each testedglass disk. The measured inhibition zone diameters (minus thediameter of the glass disk,12 mm) are given in Table 1. It can be seenfrom Figs. 8 and 9 that all tested glasses (even silver free glasses)demonstrated different antibacterial effects against the testedmicro-organisms as indicated by the clear zone around each glassdisk. Figs. 8 and 9 also show that the glass antibacterial effectdepends on the glass composition, Ag2O content and type of thetested micro-organism. S. aureus was found to be the most suscep-tible micro-organism to the tested antibacterial glasses. The degreeof susceptibility of the tested micro-organisms to the tested anti-bacterial glasses was in this order S. aureus > P. aeruginosa > E. coli.Figs. 10e12 show the variation of the inhibition zone diameterwith the Ag2O content in the all studied glasses. As displayed inFigs. 10e12, a gradual increase in the inhibition zone diameter (theantibacterial activity is proportional to the size of inhibition zone)

was seenwith increasing Ag2O content in the glass. The biggest zoneof inhibition among silver-doped glasses was observed for thehighest silver releasing glass, namely H3Ag2 against S. aureusmicro-organism. Figs. 13e15 show the variation of the inhibition zonediameter with dissolution rate for silver free glasses. The biggestzone of inhibition among silver free glasses was observed for theglass with the highest dissolution rate, namely H3 against S. aureusmicro-organism as shown in Figs. 13e15.

4. Discussion

4.1. Glass forming region

A glass forming region which contains � 60 mol% of P2O5 wasobserved in the quaternary system P2O5eCaOeNa2Oex Ag2O,x ¼ 0.5, 1 and 2 mol%, whereas compositions containing � 55 mol%of P2O5 showed metallic silver particles precipitation. TheP2O5eCaOeNa2OeAg2O glasses were prepared under normalmelting conditions without any special precautions.

Since silver is a noble element, its oxide is easily reduced at hightemperatures. The decomposition temperature of Ag2O underambient conditions is calculated from its thermodynamic data to be421 K. It is well known that silver may exist in glass in one or morethan one of its common states (Ag0, Agþ or Ag2þ). The solubility ofAg2O in glass melts is an important factor for effective production ofAg2O containing glasses as silver is known to have different degrees

Fig. 8. Photos of Petri dishes after conducting agar disk-diffusion assays at 37 �C for 24 h with (a) S. aureus, (b) P. aeruginosa and (c) E. coli as test micro-organisms. (For base glasses).

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992986

of solubility in glass melts. In phosphate melts, silver has goodsolubility.

The Ag2O added to the glass batch over its solubility is reducedto metallic silver and it is of no use to increase Ag2O content in theglass beyond its solubility limit. It is known that the solubility ofAg2O in glass melts is limited by thermodynamic factors and ishighly dependent on the glass composition, temperature duringmelting, and the oxygen potential [13,14]. Therefore the glasscomposition should be selected carefully because it is difficult tocontrol the oxygen potential during the melting process.

4.2. Structure of studied glasses

It is well known that the phosphate network structure consistsof a series of PO4 tetrahedral units connected by three bridgingoxygens. The network connectivity can be described in terms ofa Qn distribution as shown, where n is the number of bridgingoxygens. Q0 corresponds to isolated tetrahedra (orthophosphategroups), Q1 to end groups (pyrophosphate), Q2 to middle groups(metaphosphate) and Q3 to branching groups (ultraphosphate).When a modifier oxide is added, disruption of the main networkoccurs with the modifier oxide cations occupying interstitial

positions. It is accepted that the structure of ultraphosphates,metaphosphates, and polyphosphates are dominated by (Q3 & Q2),(Q2), and (Q2 & Q1) units, respectively.

The FT-IR spectra of some undoped and Ag2O-dopedP2O5eCaOeNa2O glasses shown in Fig. 5 are typical of phosphate-based glasses showing the characteristic absorption bands of PO4groups.

The FT-IR spectra of these glasses appear to be dominated bymetaphosphate (Q2) and ultraphosphate structural units (Q3).Three main spectral regions can be distinguished in the FT-IRspectra of the glasses as follows: 400e800 cm�1, 800e1200 cm�1

and 1200e1400 cm�1. In the spectral region ( 400e800 cm�1), theband at w 450e470 cm�1 and the shoulder at w 530 cm�1 may beattributed to harmonics of bending vibrations of OePeO and O]PeO linkages.

Another band in the frequency region (400e800 cm�1) is theabsorption band at about 720e750 cm�1. This band may beattributed to the symmetric stretching vibrations of the PeOePlinkages, ys (PeOeP) modes [15]. This absorption band at about750 cm�1 shifts towards lower frequency with increasing theconcentration of Ag2O. The variation of the frequency of PeOePbonds with increasing Ag2O content is consistent with breakage of

Fig. 9. Photos of Petri dishes after conducting agar disk-diffusion assays at 37 �C for 24 h with (a) S. aureus, (b) P. aeruginosa and (c) E. coli as test micro-organisms. (For glassescontaining 2 mol % Ag2O).

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992 987

cyclic PeOeP bonds in the glass when the Ag2O acts as a networkmodifier. The FT-IR absorptions in the spectral region800e1200 cm�1 was found to be sensitive for the different meta-phosphate groups in the form of chain-, ring-, terminal groups[16,17]. This area in the FT-IR spectra of all the studied glasses ischaracterized by the presence of four absorption bands at about910e920, 1017e1050, w1100 and 1160e1170 cm�1. The absorptionband at w 910 cm�1 is assigned to the asymmetric stretchingvibrations of PeOeP groups linked with linear metaphosphatechain [15]. Phosphate-based glasses with a metaphosphate struc-ture can have a chain and/or ring structure. The occurrence of theyas (PeOeP) at around 900 cm�1 in all the samples studied is anindication of a chain structure, as the analogous vibration occurs ataround 1000 cm�1 in ring structures [18]. The shift of PeOePasymmetric stretching vibration around 900 cm�1 to higher wavenumbers [19], always indicates increase of covalency proportion ofthe PeOeP bond and strengthening of glass structure withimproved chemical durability. The weak absorption band at1017e1050 cm�1 is attributed to a normal vibration mode (y3symmetric stretching) of the PO4

3- groups [20,21]. The position ofthis band is constant and does not change by the addition of varying

concentrations of the Ag2O. The band at w 1100 cm�1 can beassociated with an overlap of several modes like stretching of thePO3 terminal and PO2 middle groups. In addition, the y3 (F2) modeof the ortho anion can also contribute to the absorption at1100 cm�1 [15,22]. The band at w 1100 cm�1 was also assigned toasymmetric stretching mode of chain-terminating Q1 groups, yas(PO3)2-. The shoulder at about 1160e1170 cm�1 may be attributedto symmetric stretching mode of the two non-bridging oxygenatoms in the Q2 tetrahedral sites, ys (PO2)� mode in metaphosphategroups [23].

The spectral region 1200e1400 cm�1 showed a strong broadabsorption band in the range 1298e1330 cm�1. The appearance ofthis band (at w1300 cm�1) could be due to the presence of a frac-tion of Q3 tetrahedral units which are the most characteristicfeature of an ultraphosphate glass. This band appeared in all the FT-IR spectra of the studied phosphate-based glasses and it is attrib-uted to the asymmetric stretching of the doubly bonded oxygen, yas(P]O) modes, which are only present in the glasses with a P2O5content > 50 mol% [15,16]. Osaka et al. [24] have shown that PO4units have two bridging oxygen bonds alongwith two non-bridgingbonds such as P]O and PeO�, which are in resonance with each

Fig. 11. Variation of IZD with Ag2O content for (70ex) P2O5e30CaOex Ag2O glasses,x ¼ 0, 0.5, 1 and 2 mol%.

Table 1Glass compositions, Dissolution Rate, Density and the measured inhibition zonediameters (minus the diameter of the glass disk, 12 mm).

GlassComposition

Glasscode No.

Density D.R(g cm�2 h�1)*10�4

Inhibition zonediameter (mm)

P2O5 CaO Na2O S.aureus

P.aeruginosa

E. coli

70 20 10 G5 2.3937 3.91 27 25 2270 20 9.5 G5Ag0.5 2.4186 3.52 31 28 2570 20 9 G5Ag1 2.4415 3.12 34 31 2870 20 8 G5Ag2 2.4827 2.55 38 35 3270 30 0 G7 2.4098 3.24 25 22 1969.5 30 0 G7Ag0.5 2.4302 2.95 29 28 2369 30 0 G7Ag1 2.4523 2.67 32 30 2668 30 0 G7Ag2 2.4910 2.19 36 33 2965 10 25 H3 2.4229 6.46 30 28 2565 10 24.5 H3Ag0.5 2.4440 5.85 34 31 2965 10 24 H3Ag1 2.4647 5.32 37 35 3365 10 23 H3Ag2 2.5066 4.36 41 38 36

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992988

other. Therefore, the FT-IR spectra, as in the present study, are splitinto two bands with a higher energy and strong double bondcharacter (1298e1330 cm�1) and a lower energy band(w1100 cm�1). This means that the P]O double bond is morestrongly localized in the central position of phosphate groups. Thisassumption is confirmed by the larger relative intensity of the1298e1330 cm�1 envelope compared with that of the band at w1100 cm�1.

The weak absorption band at around 1615e1644 cm�1 observedin all spectra of phosphate glasses prepared in this work is attrib-uted to the bending vibrationmodes of OeH groups, d (H2O) modes[21]. This band demonstrates the presence of small amounts ofwater absorbed from air moisture (due to the hygroscopic characterof the phosphate glasses) during the preparation of the KBr pelletsfor infrared measurements [16]. The band positions of variousphosphate structural groups observed for the studied glass sampleswere found to be well within the ranges reported in the literaturefor different phosphate-based glasses. The positions and assign-ments [15, 19, and 23] of the IR absorption bands observed in theFT-IR spectra of the studied glasses are summarized in Table 2.

As the P2O5 content increases the band around 1320 cm�1

become broader and this is typically for ultraphosphate glasses.Also it can be seen from the FT-IR spectra of the base glasses thatthe absorption bands at about w 910 cm�1 and 720 cm�1 assigned

Fig. 10. Variation of IZD with Ag2O content for 70P2O5e20CaOe(10ex ) Na2Oex Ag2Oglasses, x ¼ 0, 0.5, 1 and 2 mol%.

to the yas (PeOeP) and ys (PeOeP) modes shift to higherfrequencies as Na2O is replaced by CaO showing an intensifiedPeOeP bonds in the glass structure and showing that the covalencyproportion in the bond of the metal ions with the non-bridgingoxygen on the [PO4] tetrahedron units increased probably becauseof the increased polarizability of the calcium ions which hasa smaller electrovalency in the Ca..O bond than in the Na..O bond.This explains the improvement in chemical durability withincreasing CaO content [19]. The band shift can be explained by theincrease in covalent character of PeOeP bonds, indicating that thePeOeP bonds are strengthened as the Na2O is substituted by CaO.This explains the improvement in the chemical durability of theseglasses as the CaO content increases.

Overall, the FT-IR spectra of the base P2O5eCaOeNa2O showedthat the shift of yas (PeOeP) and ys (PeOeP) to higher frequencies asCaO replaces Na2O is witnessed with improved chemical durabilityof P2O5eCaOeNa2O glasses and this was explained in terms of:

(i) Strengthened Ca..O bond formed by the non-bridging oxygenwith the Ca ion introduced on the [PO4] tetrahedron;

(ii) Strengthened PeOeP bond formed by the bridging oxygenbetween two [PO4] tetrahedrons.

Fig. 12. Variation of IZD with Ag2O content for 65P2O5e10CaOe(25ex) Na2Oex Ag2Oglasses, x ¼ 0, 0.5, 1 and 2 mol%.

Fig. 15. Variation of IZD with D.R for some P2O5eCaOeNa2O glasses in case of E. colimicro-organism.

Fig. 13. Variation of IZD with D.R for some P2O5eCaOeNa2O glasses in case of S. aureusmicro-organism.

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992 989

According to the dissolution mechanism of the phosphateglasses [25], these two factors determine the dissolution rate of thephosphate glasses by both limiting the velocity of hydration reac-tion and the network breakage reaction, respectively.

4.3. UVeVIS spectra

It is well known [26,27] that the optical properties of silveratoms are entirely determined in the visible region by their freeelectrons (Ag: [Kr] 4d10 5s1), while that of silver ions are deter-mined in the ultraviolet region (Agþ: [Kr] 4d10 5s0). Thus absorp-tions in the UVeVIS spectral regions can help in the elucidation ofsilver states within a glass network.

In the present work, the UVeVIS absorption spectra of someundoped and silver-doped P2O5eCaOeNa2O glasses revealed noabsorption peaks in the visible region, whereas two absorptionpeaks were observed in the ultraviolet region. A peak at bout210 nm was observed for all undoped glasses as shown in Fig. 6a,b and c except for the binary glass, G7, where this peak wasobserved at about 240 nm and it was weak and broad as shown inFig. 6b. This band (either at about 210 nm or at about 240 nm) isattributed to electronic absorption given by the phosphate glassnetwork (the threshold absorption of the host glass matrix) [28].

Fig. 14. Variation of IZD with D.R for some P2O5eCaOeNa2O glasses in case of P.aeruginosa micro-organism.

A strong absorption peak at about 230 nm was observed for allsilver-doped glasses as shown in Fig. 6awhich is presumably due toAgþ ions. Ahmed and Abdallah [29] observed three absorptionbands at 305 nm, 350 nm and 420 nm in the UVeVIS spectra ofsilver-containing soda-lime silica glass prepared by the ion-exchange process. They assigned these bands to silver ions Agþ,elemental silver Ag0 and silver crystallites (Ag0)n (n is the numberof silver atoms forming the crystallite) respectively. Paje et al. [30]observed a broad absorption band centered at w260 nm fora silver-doped silicate glass prepared by the melt-quenchingtechnique. However, they also observed the same peak in the hostglass, so they did not attribute it to silver. Borsella et al. [31] per-formed UVeVIS spectroscopic investigations on a range of silverconcentrations in soda-lime glass prepared by the ion-exchangeprocess. In contrast to Paje et al. [30] they observed an absorptionband at about 268 nm for silver ions in the soda-lime glass. Theabsorption was ascribed to transitions involving the promotion ofan electron from the 4d10 ground state to the 4d9 5s1 state. Jimenezet al. [32] observed an absorption band centered at about 275 nmfor silver-doped aluminophosphate glass and they ascribed it tosilver ions. It was reported [28] that besides the pronounced silverplasmon resonances that appear between 400 and 500 nm, theelectronic transitions involving the Agþ ion give rise to absorption

Table 2Assignment of IR absorption bands of studied glasses.

Bandfrequency (cm�1)

Assignment

450e470 Bending vibrations of OePeO bonds, d (PO2) modes.530 Fundamental bending vibrations of O]PeO bonds,

d (O]PeO) modes.660 Bending mode of PO4 units.720e750 Symmetric stretching vibrations of PeOeP linkages,

ys (PeOeP).900e920 Asymmetric stretching vibrations of PeOeP linkages,

yas (PeOeP) linked with linear metaphosphate chain.1020e1050 Vibration mode (y3 symmetric stretching) of the PO4

3-

groups.1100 Asymmetric stretching mode of chain-terminating

Q1 groups, yas (PO3)2-.1170 Symmetric stretching mode of the tow non-bridging

oxygen atoms in the Q2 tetrahedral sites, ys (PO2)�

mode in metaphosphate groups Q2.1300e1330 Asymmetric stretching of doubly bonded oxygen,

yas (P]O) modes.1610e1640 Bending vibration modes of OeH groups in HeOeH,

d (H2O) modes.

Table 3pH values of studied glasses in different times.

Glass Code No. pH

1h 2h 3h 4h 5h 6h

G5 2.98 2.87 2.81 2.72 2.65 2.54G5Ag0.5 3.05 2.93 2.87 2.79 2.71 2.63G5Ag1 3.14 3.06 2.98 2.90 2.81 2.74G5Ag2 3.30 3.21 3.15 3.04 2.95 2.85G7 3.07 2.95 2.88 2.80 2.73 2.62G7Ag0.5 3.12 3.05 2.96 2.89 2.80 2.69G7Ag1 3.25 3.13 3.03 2.95 2.87 2.77G7Ag2 3.41 3.32 3.23 3.13 3.01 2.92H3 2.84 2.75 2.68 2.60 2.52 2.41H3Ag0.5 2.91 2.82 2.75 2.67 2.58 2.50H3Ag1 3.01 2.93 2.85 2.77 2.68 2.59H3Ag2 3.20 3.11 3.01 2.94 2.85 2.76

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992990

bands located between 200 and 230 nm, whereas the electronictransitions of metallic Ago atoms appear in the 250e330 nmspectral range.

Accordingly, it is reasonable to ascribe the absorption band ataround 235 nmwhich appeared in the UVeVIS spectra of the Ag2O-doped P2O5eCaOeNa2O glasses to electronic transitions involvingAgþ ions, [4d10 / 4d9 5 s1]. Although these transitions are parity-forbidden, they are partially allowed in a solid due to vibrationalcoupling. The observed red shift of the absorption edge (the shiftfrom 230 to 245 nm) associated with increasing Ag2O content,appears to be related to several effects. The progressive introduc-tion of silver oxide is responsible for the red shift of the absorptionedge. Regarding silver and its role as a network modifying oxideimplies that the red shift (caused by Ag2O addition) is related to theincorporation of silver ions in glass structure. Also this shift may bearise from an increase in the concentration of non-bridgingoxygens with increasing silver content or may be due to thedecrease of the smallest AgeO distance existing in the family ofsilver sites.

4.4. Antibacterial activity

An antibacterial activity of a substance is an indication of itsability to kill bacteria or inhibit their growth. In the present study,the tested silver free and silver-doped phosphate-based glassesshowed different antibacterial effects against S. aureus, P. aerugi-nosa and E. coli micro-organisms as shown in Figs. 8 and 9. Theantibacterial effect is indicated by the clear zone (zone of nobacterial growth) around each glass disk. As shown in these figures,the silver-doped glasses demonstrated more potent antibacterialeffects than the silver free glasses. According to the results of theagar disk-diffusion assays, all glass compositions examined wereable to inhibit the growth of Gþ and G- bacteria. The antibacterialeffect as shown in Figs. 8 and 9 varies according to the glasscomposition, Ag2O contents and the type of the micro-organism.The degree of susceptibility of the tested micro-organisms to thetested antibacterial glasses was in this order S. aureus > P.aeruginosa > E. coli. The explanations of the antibacterial effects ofsilver free and silver-doped glasses are given in the followingsections:

4.4.1. Antibacterial activity of silver free glassesIn the present work, some P2O5eCaOeNa2O glasses were used

as control samples in the antibacterial activity test. It was thoughtthat these silver free glasses would not show any antibacterialeffects against the tested micro-organisms. Interestingly, thesesilver free glasses namely, G5, G7, and H3 produced inhibitory zonesof different sizes (Fig. 8).

This antibacterial effect of these silver free glasses can beexplained in terms of the glass composition, the glass dissolutionrate, and the pH changes of the medium. Valappil et al. [33]observed a zone of inhibition for S. aureus, MRSA and C.difficile intesting the antibacterial activity of gallium free45P2O5e16CaOe39Na2O glass. They attributed the antibacterialeffect of this gallium free glass to the change in pH during the glassdegradation. Pickup et al. [23] investigated the antibacterial activ-ities of a Ga-doped solegel phosphate-based glasses of composition(CaO)0.30 (Na2O)0.20�x (Ga2O3)x (P2O5)0.50 where x ¼ 0 and 0.03 mol%. They observed a small zone of inhibition (7 mm) for the galliumfree glass and they attributed it to either a change in pH as the glassdissolves or by reduced water activity as ions leach out.

The glass compositions investigated contain 65 and 70 mol% ofP2O5 and the remainder is made up of CaO and Na2O. Accordingly,due to their high P2O5 contents, these glasses have acidic compo-sition and this means that the dissolution of such glasses changes

the pH and produces acidic media (since more acidic phosphateions are released).

The pH value of the medium (in which the glass dissolves) andinhibition nation zone diameter were found to be related to theglass dissolution rate (Table 1 and Figs. 13e15). The dissolution rateof P2O5eCaOeNa2O glasses depends on the contents of P2O5, CaOand Na2O in the glass. From the results of dissolution of suchglasses, it was found that the glass dissolution rate is directlyproportional to either P2O5 or Na2O contents and inverselyproportional to CaO content. The glass dissolution study showedthat the dissolution rate of these glasses as shown in Table 1 andFigs. 13e15 are in this order H3 > G5 > G7 i.e. H3 has the fastestdissolution rate, whereas G7 has the slowest dissolution rate in thisglass series. As shown in Tables 1 and 3 a decrease in the pH wasseen with the increases in the dissolution rate since more acidicphosphate ions are released.

It is well known that the acidity or alkalinity of the mediumaffects the growth of bacteria. The pH affects the rate of enzymeaction and plays a role in determining the ability of bacteria to growor survive in particular environments. Most bacteria survive nearneutral conditions and grow optimally within a narrow range of pHbetween 6.7 and 7.5. Since the optimal pH of all tested bacteria inthis study is close to neutral. Thus, the decrease in pH during thedissolution of P2O5eCaOeNa2O glasses could explain the bacterialgrowth inhibition produced by such glasses. The more the phos-phate ions released the lower is the pH and the greater the anti-bacterial effect. Thus the low pH produced by glass dissolution wascertainly a critical factor for glass antibacterial effect. Overall, theantibacterial effect of silver free glasses was influenced by glasscomposition, glass dissolution rate and the dissolution conditionsin the glass surroundings. This might explain at least some of thedifferences in the antibacterial action of glass with varying chem-ical compositions. The antibacterial activity of silver free glasseswas closely related to the dissolution rate of the glasses becausehigh dissolution rates cause a decrease of pH or an increase of theion concentration of the media, and this resulted in antibacterialactivity. Therefore, the results of silver free glass dissolution, theirpH changes during dissolution and their antibacterial effectscorrelate well with each other. The antibacterial effect of the silverfree glasses may be attributed to other factors [34e36] e.g. the ionicstrength of the medium where high concentrations of calcium,sodium and phosphates ions likely to be released from the glassduring dissolution could cause perturbations of the membranepotential of bacteria and such as osmotic effects caused by thenonphysiological concentration of ions such as sodium, calcium,and phosphate ions dissolved from the glass and lead to change theosmotic pressure in the vicinity of the glass.

The exact mechanism of the antibacterial action of these silverfree glasses is unknown. Therefore, we conclude that the

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992 991

antibacterial effect observed with these glasses can be explained bythe dramatic changes in the physicochemical characteristics of theculture medium (pH, ionic strength, and osmotic pressure) whichoccur as a consequence of the glasses dissolution. Thus the anti-bacterial action of glass is influenced by its chemical compositionand the dissolution conditions in its surroundings.

4.4.2. Antibacterial activity of silver-doped glassesIn the present work, the addition of Ag2O to P2O5eCaOeNa2O or

P2O5eCaO glasses has been found to potentate its antibacterialactivity. Silver-doped glasses showed increased antibacterialactivities (depending upon the Ag2O content) more than silver freeglasses as shown in Figs. 8 and 9 and as indicated by the measuredinhibition zone diameters Table 1 and Figs. 10e12. This increase isattributed to the release of the Agþ ions which are well known asantibacterial metal ions [37].

An increase in the antibacterial activity as represented by theincrease in the inhibition zone diameter was seen with increasingthe Ag2O content in the glass as displayed in Figs.10e12. This was ina good agreement with the results of Agþ ions release seen inwaterwhere an increase in concentrations of silver ions released fromglass into water was seen with the increase in Ag2O content.

Generally, in an aqueous medium or in presence of moisture, thesilver-doped glass gradually dissolves depending on its dissolutionrate and during its dissolution, the silver ions (the antibacterialactive agents) incorporated into its structure are released into themedium and inhibit the growth of bacteria. Thus the mechanismfor antibacterial action of silver-doped glasses is bacterial growthinhibition by the silver ions released from the glass. Bellantoneet al. [38] investigated the antibacterial effects of Ag2O-dopedbioactive glasses on S. aureus, E. coli, and P. aeruginosa. The anti-bacterial action of the silver-doped bioactive glass was attributed tothe leaching out of Agþ ions from the glass matrix. Kim et al. [39]investigated the antimicrobial effects of various ceramics againstE. coli using viable count and growth rate studies. They concludedthat Agþ ions interfered with the metabolism of the micro-organism, thus inhibiting its growth.

The exact mechanism of antibacterial action of silver ions is stillunknown. Antibacterial mechanisms of silver ions might differaccording to the species of bacteria. Generally, most antibacterialagents exert their antibacterial action by four principal modes ofaction. These modes include: inhibition of bacterial cell wallsynthesis, inhibition of protein synthesis, inhibition of synthesis ofbacterial RNA and DNA or inhibition of a metabolic pathway. Inbacteria, silver ions are known to react with bacterial nucleophilicamino acid residues in proteins, and attach to sulphydryl, amino,imidazole, phosphate, and carboxyl groups of membranes orenzymes, resulting in protein denaturation [40]. Silver is alsoknown to inhibit a number of oxidative enzymes such as yeastalcohol dehydrogenase [41] the uptake of succinate by membranevesicles [42] and the respiratory chain of E. coli, as well as causingmetabolite efflux [43] and interfering with DNA replication. Holtand Bard [44] examined the interaction of silver ions with therespiratory chain of E. coli. They found that an addition of �10 mMAgNO3 to suspended or immobilized E. coli resulted in stimulatedrespiration before death, signifying uncoupling of respiratorycontrol from ATP synthesis. This was a symptom of the interactionof Agþ with enzymes of the respiratory chain. Feng et al. [4] studiedthe antibacterial effect of silver ions on E. coli and S. aureus andsuggested that the antibacterial mechanism was due to DNA notbeing able to replicate, and proteins becoming inactivated aftercontact with Agþ ions.

One of the primary targets of Agþ ions, specifically at lowconcentrations, appears to be the Naþ-translocating NADH:ubiquinone oxidoreductase system [45,46]. Silver has also been

shown to be associated with the cell wall [47], cytoplasm and thecell envelope [48]. Chappell and Greville [49] acknowledged thatlow levels of Agþ ions collapsed the proton motive force on themembrane of bacteria, and this was reinforced by Mitchell’swork [50]. Dibrov et al. [51] showed that low concentrations ofAgþ ions induced a massive proton leakage through the bacterialmembrane, resulting in complete de-energization and, ultimately,cell death. Overall, there is consensus that surface binding anddamage to membrane function are the most important mecha-nisms for the killing of bacteria by Agþ ions. The results of anti-bacterial activity showed that the silver free and silver-dopedglasses exhibited different antibacterial effects against the testedbacteria and the sensitivity of G� and Gþ bacteria to the anti-bacterial glasses was different. A remarkable difference was seenbetween G- bacterium (E. coli) and Gþ bacterium (S. aureus),while a small difference was found between the two G- bacteria. Itis known that the cell wall of G- bacteria is composed of highproportion of phospholipids, lipopolysaccharides and proteins(the cell wall of G- bacterium is chemically more complex thanthat of Gþ bacteria), whereas peptidoglycan is the majorcomponent of the cell wall of Gþ bacteria. This fact would possiblycontribute to the difference of antibacterial effects between G�and Gþ bacteria.

5. Conclusion

1- A glass forming region containing � 60 mol% of P2O5 wasobserved in the quaternary P2O5eCaOeNa2Oex Ag2O system,x ¼ 0.5, 1 and 2 mol%.

2- Generally, all the prepared silver free and silver-doped glasseswere (as observed visually) homogeneous, free from solidinclusions, transparent and colorless. For 2 mol% silver-dopedP2O5eCaOeNa2O glasses, the XRD patterns indicated theabsence of formation of any crystalline phases and thus XRDensured the amorphous nature of these glasses.

3- Measurements of pH changes during dissolution of silver-freeand silver-doped glasses in water revealed a decrease of waterpH with increasing time of glass dissolution. It was found thatthe magnitude of the pH drop increases with the increase inglass dissolution rate.

4- Silver-doped glasses showed less pH drop than silver-freeglasses. In agar disk-diffusion assays, all the tested silver-freeand silver-doped glasses demonstrated different antibacterialeffects (depending on the glass composition and the type of thetested micro-organism) against S. aureus, P. aeruginosa and E.coli micro-organisms as indicated by the clear zone (zone of nobacterial growth) around each tested glass disk.

5- For silver-free glasses an increase in bacterial growth inhibitionzone diameter was observed with the increase in the glassdissolution rate and with the decrease in pH, whereas for silver-doped glasses an increase in bacterial growth inhibition zonediameter was observed with increasing Ag2O content.

6- S. aureus as a Gþ bacterium was found to be the most suscep-tible micro-organism to the tested antibacterial glasses. Thedegree of susceptibility of the tested micro-organisms to thetested antibacterial glasses was found in this order S. aureus > P.aeruginosa > E. coli.

7- There is a good agreement between the antibacterial activities ofsilver-free glasses, their dissolution rates and their pH changesduring their dissolution. Also there is a good agreementbetween the antibacterial activities of silver-doped glasses andthe concentrations of silver ions released into water during theirdissolution.

A.A. Ahmed et al. / Solid State Sciences 13 (2011) 981e992992

References

[1] J. Clement, J.M. Manero, J.A. Planell, G. Avila, S. Martinez, J. Mater. Sci. Mater.Med. 10 (1999) 729.

[2] I. Ahmed, M.P. Lewis, I. Olsen, J.C. Knowles, Biomaterials 25 (3) (2004) 491.[3] K. Franks, I. Abrahams, G. Georgiou, J.C. Knowles, Biomaterials 22 (2001) 497.[4] Q.L. Feng, J. Wu, G.Q. Chen, F.Z. Cui, T.N. Kim, J.O. Kim, J. Biomed. Mater. Res. 52

(2000) 662e668.[5] M. Kawashita, S. Tsuneyama, F. Miyaji, T. Kokubo, H. Kozuka, K. Yamamoto,

Biomaterials 21 (2000) 393e398.[6] A.P. Adams, E.M. Santschi, M.A. Mellencamp, Vet. Surg. 28 (1999) 219e225.[7] Y. He, D.E. Day, Glass Technol. 33 (1992) 214.[8] A.E. Marino, S.R. Arrasmith, L.L. Gregg, S.D. Jacobs, G. Chen, Y. Duc, J. Non-Cryst

Solids 289 (2001) 37.[9] C.F. Drake, Pest Sci. 9 (1978) 441.

[10] C.F. Drake, W.M. Allen, Biochem. Soc. Trans. 13 (1985) 520.[11] A.M. Mulligan, M. Wilson, J.C. Knowles, J. Biomed. Mater. Res. 67A (2003)

401e412.[12] I. Ahmed, D. Ready, M. Wilson, J.C. Knowles, J. Biomed. Mater. Res. 79A (2006)

618e626.[13] D.A. McKeown, H. Gan, I.L. Pegg, J. Non-Cryst. Solids 351 (2005) 3826e3833.[14] M. Bosca, L. Pop, G. Borodi, P. Pascuta, E. Culea, J. Alloys Compounds 479

(2009) 579e582.[15] R.C. Lucacel, A.O. Hulpus, V. Simon, I. Ardelean, J. Non-Cryst. Solids 355 (2009)

425e429.[16] Y.M. Moustafa, K. El-Egili, J. Non-Cryst. Solids 240 (1998) 144.[17] E.I. Kamitsos, A.P. Patsis, M.A. Karakassides, G.D. Chryssikos, J. Non-Cryst.

Solids 126 (1990) 52.[18] A. Rulmont, R. Cahay, M.L. Duyckaerts, P. Tarte, Eur. J. Solid State Inorg. Chem.

28 (1991) 207.[19] H. Gao, T. Tan, D. Wang, J. Control. Release. 96 (2004) 21e28.[20] A.H. Khafagy, M.A. Ewaida, A.A. Higazy, M.M.S. Ghoneim, I.Z. Hager,

R.R. El-Bahnasawy, J. Mater. Sci. 27 (1992) 1435.[21] S.S. Das, B.P. Baranwal, P. Singh, V. Srivastava, Prog. Crystal Growrh Characr 45

(2002) 89e96.[22] D. Ilieva, B. Jivov, G. Bogacev, C. Petkov, I. Penkov, Y. Dimitriev, J. Non-Cryst.

Solids 238 (2001) 195.[23] D.M. Pickup, S.P. Valappil, R.M. Moss, H.L. Twyman, P. Guerry, M.E. Smith,

M. Wilson, J.C. Knowles, R.J. Newport, J. Mater. Sci. 44 (2009) 1858e1867.

[24] A. Osaka, K. Takahashi, M. Ikeda, J. Mater. Sci. Lett. 3 (1984) 36.[25] B. C Bunker, G.W. Arnold, J.A. Wilder, J. Non-Cryst. Solids 64 (1984) 291.[26] P. Chakraborty, J. Mater. Sci. 33 (1998) 2235.[27] R.H. Doremus, J. Chem. Phys. 42 (1965) 414.[28] J. Lu, J.J. Bravo, A. Takahashi, M. Haruta, S.T. Oyama, J. Catal. 232 (2005) 85.[29] A.A. Ahmed, E.W. Abd Allah, J. Am. Ceram. Soc. 78 (10) (1995) 2777e2784.[30] S.E. Paje, J. Llopis, M.A. Villegas, J.M. Fernandez, Appl. Phys. A 63 (1996) 431.[31] E. Borsella, G. Battaglin, M.A. García, F. Gonella, P. Mazzoldi, R. Polloni,

A. Quaranta, Appl. Phys. A 71 (2000) 125.[32] J.A. Jimenez, S. Lysenko, G. Zhang, H. Liu, J. Electronic Mater. 36 (2007).[33] S.P. Valappil, D. Ready, E.A. Abou Neel, D.M. Pickup, W. Chrzanowski,

L.A. O’Dell, R.J. Newport, M.E. Smith, M. Wilson, J.C. Knowles, Adv. Funct.Mater. 18 (2008) 732e741.

[34] P. Stoor, E. Söderling, J.I. Salonen, Acta Odontol Scand. 56 (1998) 161e165.[35] E. Munukka, O. Lepparanta, M. Korkeamaki, M. Vaahtio, T. Peltola, D. Zhang,

L. Hupa, H. Ylanen, J.I. Salonen, M.K. Viljanen, E. Eerola, J. Mater. Sci. Mater.Med. 19 (2008) 27e32.

[36] O. Lepparanta, M. Vaahtio, T. Peltola, D. Zhang, L. Hupa, M. Hupa, H. Ylänen,J.I. Salonen, M.I.K. Viljanen, E. Eerola, J. Mater. Sci. Mater. Med. 19 (2008)547e551.

[37] A.B.G. Lansdown, J. Wound Care 11 (4) (2002) 125.[38] M. Bellantone, H.D. Williams, L.L. Hench, Antimicrobial. Agents. Chemother.

46 (2002) 1940e1945.[39] T.N. Kim, Q.L. Feng, J.O. Kim, J. Wu, H. Wang, G.C. Chen, F.Z. Cui, J. Mater. Sci.

Mater. Med. 9 (1998) 129e134.[40] S.L. Percival, P.G. Bowler, D. Russell, J. Hosp. Infect. 60 (2005) 1e7.[41] P.J. Snodgrass, B.I. Vallee, F.L. Hoch, J. Biol. Chem. 235 (1960) 504e508.[42] M.K. Rayman, T.C.Y. Lo, B.D. Sanwal, J. Biol. Chem. 247 (1972) 6332e6339.[43] W.J.A. Schreurs, H. Rosenberg, J. Bacteriol. 152 (1982) 7e13.[44] K.B. Holt, A.J. Bard, Biochemistry 44 (2005) 13214e13223.[45] A.L. Semeykina, V.P. Skulachev, FEBS Lett. 269 (1990) 69e72.[46] M. Hayashi, T. Miyoshi, M. Sato, T. Unemoto, Biochim. Biophys. Acta 1099

(1992) 145e151.[47] H.S. Rosenkranz, H.S. Carr, Antimicrob. Chemother. 2 (1972) 367e372.[48] P.A. Goddard, A.T. Bull, Appl. Microbiol. Biotechnol. 31 (1989) 314.[49] J.B. Chappell, G.D. Greville, Nature 174 (1954) 930e931.[50] P. Mitchell, Biol. Rev. 41 (1966) 445e502.[51] P. Dibrov, J. Dzioba, K.K. Gosink, C. Hase, Antimicrob. Agents Chemother.

8 (2002) 2668e2670.